How to handle cofound variables? [D]
A developer is concerned about a radar model learning environmental artifacts rather than class distribution. The issue stems from a radar characteristic where farther objects return fewer points, potentially leading the model to associate larger ranges with larger objects. Despite improved performance and no classical overfitting, the developer fears the model is learning this confounding variable, impacting its accuracy.
Time & source
Times shown in UTC
Display time zone: UTC
Local time zone unavailable; showing UTC.
PublishedOffset at this time: UTC+0Sep 11, 2026, 18:57 UTC
IngestedOffset at this time: UTC+0Sep 12, 2026, 10:01 UTC
- Published
- Sep 11, 2026, 18:57
- Ingested
- Sep 12, 2026, 10:01
- Source type
- Dev community
- Tier
- Community
- Source status
- Sync delayed
Tier is a per-source editorial setting, not a per-item score.
Discussion trend
The percentage is based on collected discussion signal, not new comments or independent people. The curve only compares the same topic across time.
edit: confound
Hello all,
I am working on a object classification with a automotive radar point clouds. I compared many models and feature vectors.
Once i used range as feature, all models scored higher f1 in all K validation sets and on the final test set.
One particular artifact of a radar, is that as the farther the object is the less number of points it returns to the radar. Although the performance improved and there is no overfit in the classical sense, i am afraid my model is learning the environment not the class distribuiton and even worse, its learning that big range means big object.
How can i stress test this claim? Should i try to split the data sets so range distribution differs? Or not even using the feature at all and accept lower performance?
Would appreciate your insights.